Industry · Software · Tech Policy · Events
Semiconductors, foundries, advanced packaging and the global silicon supply chain, covered by The Silicon Ledger.
The bottleneck in artificial intelligence infrastructure is no longer raw compute speed but rather the ability to move data efficiently through interconnected systems while managing power and cooling demands.
This week's technology industry developments reveal how artificial intelligence is reshaping semiconductor manufacturing, patent licensing and the global race for AI capability, from neon supply disruptions to new chip-stacking techniques.
According to leaked factory notices, Nvidia has halted production of GeForce RTX 5090 graphics cards to reserve the GB202 processor for higher-margin data center and professional products.
A new firmware release from Gigabyte signals Intel's plan to extend the LGA 1700 socket with fresh processors launching in early 2027, supporting both DDR4 and DDR5 memory on existing B760 and H610 boards.
Verified Q3 2026 reporting dates and call times for ASML, TSMC, Texas Instruments, Samsung, Intel, AMD and Arm, and what is still unconfirmed for Nvidia, SK hynix, Qualcomm and Micron.
The TeamGroup T-Force G50 Evo delivers PCIe 4.0 speeds at 10 cents per gigabyte, marking one of the lowest prices for a high-capacity SSD this shopping season.
Construction equipment makers are turning to physical AI to address labor shortages and productivity challenges, but the technology demands fundamentally different data center capabilities than traditional AI training.
NetApp is positioning itself as more than a storage vendor, arguing that enterprises can make their existing data AI-ready by keeping it in place rather than rebuilding their infrastructure from scratch.
